The first part is causing a struggle. I have 2 columns and another 2 columns (for the two sets of data points). A user will know the two end points of the curve, so they'd list those two data points. They might also know the middle point, so they'd add that to the data sets. Now you have 2 datasets, each with 3 points, and these can be plotted on Scatter X,Y pllots. The user would also know that they want a linear curve, quadratic, cubic, or quartic curve to be fit to those points. Trouble starts here:

*We actually want to visualize this curve as a function of 'y' data points*, i.e. x is a function of y.

**Is this possible, to begin with?**

Next after excel generates a regression formula that the user is happy with, to move to action (2), I think I need to populate a list of common data points -- for more complex functions, or where one data set doesn't quite align with the second data set, you can't simply add these together... The list of common data points would be much more populated than the initially user-input data points. A user could, of course, manually do this, by taking the regression function from the graphs and generating the longer list of data points manually. Then they simply add the X-values at the common Y-points, and everything is complete... But it's the intermediate step of taking the regression function and populating that list of regression points to a common set of Y-values. Any thoughts on this?

I also feel like this is a common problem that engineers face, so I wonder if there's a guide or a trick to doing this. i.e. Engineers often know a few characteristic points for a curve, an they know the theoretical curve is defined by a certain function (second, third, or fourth order). They want to input those points, get the curve drawn (brings us to the end of my step (1)), then maybe they want to superimpose two curves on each other (my step (2)).

Thoughts?